202 Sir David Brewster on the Radiant Spectrum. 



mains constant there in spite of the rapid rotation of the ring. 

 The least diameter of the ellipsoidal bulb which I used was 15 

 millims., and the maximum 20 millims. 



I repeated the same experiment with a helicoidal film ; the 

 bulb remained at one point of the solid helix, the entire surface 

 of which it could traverse without leaving the film when I 

 turned the axis of the surface, keeping it horizontal. The termi- 

 nation of the laminar surface on the bulb was then a skew curve. 



The following is what I imagine to be the explanation of the 

 effects which I have noticed. As soon as the plane film touches 

 the bulb, it fastens on to it in a section whose area expresses the 

 diminution of the surface of this film ; as this is subject in all 

 parts to the force of contraction of the liquid, it tends to become 

 as small as possible, which is the case when the section acb 

 (fig. 4) immersed in the plane of the ring is a maximum ; the equi- 

 librium is then stable. The phenomenon is similarly explained 

 in the case of a helicoidal film ; the surface of intersection of the 

 surfaces of the bulb and of the film is then a skew curve. 



XXVI. On the Radiant Spectrum. 

 By Sir David Brewster, K.H., F.R.S.* 



I HAVE given the name of Radiant Spectrum to a pheno- 

 menon which I discovered in 1814, and which I described 

 to this Society in the early part of that year. 



It will be understood from fig. 1, which represents the bril- 

 liant radiation which surrounds a very small image of the sun, 

 when it is formed either by reflexion or refraction, or otherwise. 



Fig\i. 



!Fig\2. 



* From the Proceedings of the Royal Society of Edinburgh, Session 

 1866-67. Communicated bv the Author. 



